The Consequences of Educational Voucher Reform in Chile
In an effort to boost student achievement and reduce income-based gaps, the Chilean government passed the Preferential School Subsidy Law (SEP) in 2008, which altered the nation’s 27-year-old universal school-voucher system dramatically. Implementation of SEP increased the value of the school voucher by 50 percent for “Priority students”, primarily those whose family incomes fell within the bottom 40 percent of the national distribution. To be eligible to accept the higher-valued vouchers from these students, schools were required to waive fees for Priority students and to participate in an accountability system.
Using national data on the mathematics achievement of 1,631,841 Chilean 4th-grade students who attended one of 8,588 schools during the year 2005 through 2012, we address two research questions (RQs):
1. Did student test scores increase and income-based score gaps become smaller during the five years after the passage of SEP?
2. Did SEP contribute to increases in student test scores and, if so, through what mechanisms?
We addressed these RQs by fitting a sequence of multi-level interrupted time-series regression models, supplemented by other descriptive analyses. We found that:
1. On average, student test scores increased markedly and income-based gaps in those scores declined by one-third in the five years after the passage of SEP.
2. The combination of increased support of schools and accountability was the critical mechanism through which the implementation of SEP increased student scores, especially in schools serving high concentrations of low-income students. Migration of low-income students from public schools to private voucher schools played a small role.
We interpret these findings as more supportive of improved student performance than other recent research on the Chilean policy reform.
Financial support for the research on which this paper is based was provided by the Interamerican Development Bank. The views expressed herein are those of the authors and do not necessarily reflect the views of the National Bureau of Economic Research.